Jump to content

more info related to buy ins oct 16th


Recommended Posts

This was in response to an analogy that stated that a bad player would be a favorite over me in a $5-$10 NL game if he started with $125,000 to my $2,000 bankroll: I would be a favorite over a worse player if I had $2000.00 in front of me and my opponent had $125,000 in front of him provided the blinds stayed small, like $5-$10. Here is the thing, I don't HAVE to continue to risk my whole bankroll against the big stack in a cash game. I could buy in for $1000 and if I lost that, I could rebuy for another $100 if necessary. If I was the better player, chances are that after a six hour session I would be ahead. Let's say for this example I won $2000 after six hours. Now I could quit, and the next day buy in for $1000 again. If I lost that, again I could buy in for another $1000. If I was the better player, and had the time to do so, I would win every last dollar of that $125,000. Also, I would be the FAVORITE to do so. I understand what you are thinking: you are thinking in a freeze out that the big stack would win more often than that. Of course, while that's true, if the $2000 stack happens to win he gets 62.5 to 1 on his money! So, if the $2000 stack was able to win even the freezeout just once out of 50 times, he'd make a very nice profit. Does that help you?

Link to post
Share on other sites

As far as chip stacks go I am not a pro and I have made soem pretty good comebacks on a shortstack so I can see how it would be perfectly conceivable that a proffessional with impecable insight would be able to make a comeback and eventually overtake the game.

Link to post
Share on other sites
Unless the guy repeatadely goes all-in.
Exactly!! If the guy with the $125,000 goes all in every single time against the guy with the $2,000 who then buys in again for another $1,000 and then another $1,000, and then manages to get his stack up to $5,000, then the guy with the $125,000 continies to go all in, all he has to do is win once to break the other guy. I think Daniel fails to see that over a long session, the guy with the $2,000 is going to bust every single time, unless he decides to play for 30 minutes and take his $5,000 and run, but these are poker players who like to play for hours and it's inevitable that the long session will turn into a bust for him. Now Daniel surely has to try and look at it this way to understand the point you are trying to make as well as others here. Daniel could be going against Chan or Doyle and if Chan started with the 2k and Daniel with the 125k and it turns into a long session, Chan can kiss his money good bye!
Link to post
Share on other sites

Even if the guy goes all in every hand, you'd still be a favorite to make money off of him. If a guy goes all in BLIND every hand and you get to look at your cards, then you will likely have a nice advantage when you go to the river. If you know he is doing that, you should be at least a 2-1 favorite when the money goes in. Again, no one says you have to KEEP doubling up your bet though! You can do it for $1000, cash out, then do it for $1000 again. Each time you bet your $1000 you will be a favorite to earn. For him to knock you off of your $2000 bankroll, he'd have to win as at least a 2-1 underdog, back to back. If you win the first one, then you'd have three bullets against him so he'd need to win three races in a row as at least a 2-1 dog. If you were able to get on a rush and get your bankroll up to $10,000, you could keep knocking him off $1000 at a time until you had all of his money.

Link to post
Share on other sites
Even if the guy goes all in every hand, you'd still be a favorite to make money off of him. If a guy goes all in BLIND every hand and you get to look at your cards, then you will likely have a nice advantage when you go to the river. If you know he is doing that, you should be at least a 2-1 favorite when the money goes in. Again, no one says you have to KEEP doubling up your bet though! You can do it for $1000, cash out, then do it for $1000 again. Each time you bet your $1000 you will be a favorite to earn. For him to knock you off of your $2000 bankroll, he'd have to win as at least a 2-1 underdog, back to back. If you win the first one, then you'd have three bullets against him so he'd need to win three races in a row as at least a 2-1 dog. If you were able to get on a rush and get your bankroll up to $10,000, you could keep knocking him off $1000 at a time until you had all of his money.
Daniel, if you cash out, don't you lose your seat at the table if others are waiting on the list? If this is the case, then using the cashout system you are referring too wouldn't be so beneficial for the guy who just built his chips up to $5,000 from $2,000, especially if he wants to continue to play.BTW: Are you playing any onliner's tonight? If so, let us know.
Link to post
Share on other sites

I think I need to post this every time I see this topic come up."I honestly have no desire to read every response made in regard to Daniel's one session at the Wynn. Having said that, I'm not sure if any of you have played with Daniel in a live game before, but I have. And I would like to talk about it a little bit, just to clear the air. I've played with Daniel in live games two times. Once was a pot limit hold em game with 20-40 blinds ( that was several years ago). And the other session was a 10-20 no limit game. Each session that we played, he had the WHOLE TABLE COVERED. He played fast and loose, and was having a good time. In fact, EVERYONE was having a good time. Both sessions, he played just about every hand. Raising, re-raising, etc. I was lucky enough to have position on him both times. The no limit session was funny, it would appear that he had a "normal" stack. But he had the 5 or 10k chip for special occassions. The first hand I played with him, he broke me. I've talked about this hand before, so I feel no need to get into it again. But I continued to play, and ended up booking a small win. The point is, he was giving action and getting action. The ONLY advantage he had was a psycholigal one. IF you want to look at that way. He also happened to be a much better player than the whole table, myself included. But because he WAS playing so wrecklessly, it evened out the playing field. And that's what many people fail to realize. If Daniel wanted to, he could sit down in a 10-20, 20-40, or 100-200 no limit game and break everyone just by playing his A game. But instead, he does everyone a FAVOR, and has a nutbar session. For those of you who have ever played with me at a "strat table" on Party, I usually do something fairly similar. Win, lose or draw, everyone has a good time. I play waaaay too many hands, and give a ridiculous amount of action. At .50-1, I feel that I can afford the loss. But you have to earn it. Obviously the same is true of Daniel, he's not giving his money away per say. But he is laying one hell of a price an you getting it. For those who don't like this, or are offended by it. The next time you see a fish tank, don't tap the glass. Until you've sat with him and played in one of these sessions, don't knock it. He certainly isn't about to quit you just because you book a 5k win. YOU as the player have the advantage. If you are afraid of going bust, don't play. If his money intimidates you psychologically, don't play. But once you get past those silly obstacles. Be sure to take advantage of people who sit down with ton of money and are willing to gamble it up with you. Statistically, you will be a favorite. Whether you're a math guy or not, you should be able to see the logic in that. Good Luck."

Link to post
Share on other sites

The Key here is (as Daniel already mentioned) The player with 2k only needs to bust the player with 125 k 1 time in 50 to show a nice profit. People who say the guy going all in has the advantage are wrong even if it is a freezeout. He only has to win a verry small percentage of the time to break even, and he should win more often than that (if the big stack is just going all in blind).

Link to post
Share on other sites

Look, say i'm a 70% favorite to win every hand, and I buy in for $1000. Daniel buys in for $125,000, and we keep playing until 1 person has all the money. Then when that happens, the person with the money cashes out, and once again I buy in for 1000$ and Daniel buys in for $125,000. Say we run this 1,000,000 times. Who do you think will end up making money? I can guarrante you to more than .01% (probably way more) that I will. I will win more than 1/20 times, and each time I win, I get 100+ times what I bought in for... anyone who is playing poker and takes the time to post on this site should realize that this is the EXACT same thing as using pot odds.In fact I should win in the long run so long as I'm over a 50% favorite each time i play a hand, it just might take more trials to work through the varience.Anyone who disagrees with me and is willing to put at least 100$ on the line should let me know =).

Link to post
Share on other sites

From mathematical perspective:Let's say you buyin for 1k and the other guy has 125k, and he uses the all-in blind strategy, that some of you seem to think is the winning play here.You'd have to double up 7 times to win all his money (1k * 2^7 = 128k) If the money would always go in when you are ~70% favorite, that means that you'd have a chance of 0.7^8=5,7%, or 1:18 to break him every time you bought in for another 1k. (even if you lost several times, you'd only have to double up 1 more time, 8 in total to take all his money, which is 1:24)So, if you bought in with 1k, you'd have to buy in approx. 20 times to be a favorite to end up with all the money, if the opponent went all in every hand. This means that in the long run, you have better chance to win 125k while only risking 18 to 24k, than the opponent has winning your 20k while risking 125k.

Link to post
Share on other sites
From mathematical perspective:Let's say you buyin for 1k and the other guy has 125k, and he uses the all-in blind strategy, that some of you seem to think is the winning play here.You'd have to double up 7 times to win all his money (1k * 2^7 = 128k) If the money would always go in when you are ~70% favorite, that means that you'd have a chance of 0.7^8=5,7%, or 1:18 to break him every time you bought in for another 1k.  (even if you lost several times, you'd only have to double up 1 more time, 8 in total to take all his money, which is 1:24)So, if you bought in with 1k, you'd have to buy in approx. 20 times to be a favorite to end up with all the money, if the opponent went all in every hand. This means that in the long run, you have better chance to win 125k while only risking 18 to 24k,  than the opponent has winning your 20k while risking 125k.
I dont think this and the prior analysis are entirely correct, and it will actually take quite a bit longer than an average of 20 buy ins to be the favorite. The reason is that if you want to be a 7/3 favorite to call the all in you are only playing pocket 8s and up, which means you are only playing 3.2% of the hands you are dealt. The blinds you lose 96.8% of the time are a huge drag on your win rate. If the blinds are 5/10, or average of 7.5, and the freeze out requires you to win 120,000, your expectation isnt 40% of all bets (70%-30%) it is dragged down to less than 1% (a guess, based on the average all in being around 30,000).
Link to post
Share on other sites
I dont think this and the prior analysis are entirely correct, and it will actually take quite a bit longer than an average of 20 buy ins to be the favorite. The reason is that if you want to be a 7/3 favorite to call the all in you are only playing pocket 8s and up, which means you are only playing 3.2% of the hands you are dealt. The blinds you lose 96.8% of the time are a huge drag on your win rate. If the blinds are 5/10, or average of 7.5, and the freeze out requires you to win 120,000, your expectation isnt 40% of all bets (70%-30%) it is dragged down to less than 1% (a guess, based on the average all in being around 30,000).
Well actually, with hands 77+, AT+ (8,4% of all hands) on average you are ~70% favorite over random hand.. So you'd only lose about $20 per every all in.Note that the 8th all in covers way more than all the blinds lost in process. Besides, the blinds lost after 3rd all in are pretty meaningless to what you are going to win when you go all in next time.However, all this is pretty meaningless because no matter how you put it, you are the favorite to win it all, if every time the money goes in you are the favorite. All the opponent's bigger stack size, and the fact thet you are trying to win it all do, is increase your variance. If you can't afford to lose 20 buyins, you are very likely to go broke in the process of trying to win it all.Of course, as dn stated, you don't have to win it all! You can quit at 20k, thus decreasing your variance quite a lot, if you don't have enough bankroll to try to win more. It's same as if you'd take a shot at higher limits than what you have bankroll for, you are more likely to go broke. Besides.. Why on earth should a $5-$10 NL player be able to win 125k in one session? If he has 20buyins as he should he can easily expect to win atleast 20k from the all in blind guy, if he quits then, without much of a risk of going broke.
Link to post
Share on other sites

Well actually, with hands 77+, AT+ (8,4% of all hands) on average you are ~70% favorite over random hand I read the scenario to be that you had to be at least a 70% favorite, not average 70%.. So you'd only lose about $20 per every all in. Not sure where you get this. Even based the 8.4% of all hands you have 9+ blinds given up for each all in or about $81. Since you then win on average only 70% of your all ins, you are losing closer to $120 per winning all in.Most importantly though, while your prior estimate based on .7 ^ 18 was conservative, when you only require an average of 70% winners you introduce a tremendous amount of variance, and a binomial distribution estimate is very inaccurate.Yes, the ultimate conclusion is that you will win against the allin strategy, but it will take far more than 20 buy ins. Lets hear some guesses as to how many, and I'll build a stochastic model to test it. My guess is 80 buy ins

Link to post
Share on other sites

I read the scenario to be that you had to be at least a 70% favorite, not average 70%If you are only going with hands that are at least 70% favorite, then you'd actually be 77% favorite on average every all in. Then the calculation would go .77^8=12,4%, because some all ins you are 70% favorite, and some even 85% (AA), which in the long run equals to .77 with hand range of 88-AA. Play with www.pokerstove.com if you can't understand this.So you'd only lose about $20 per every all in. Not sure where you get this. Even based the 8.4% of all hands you have 9+ blinds given up for each all in or about $81. Since you then win on average only 70% of your all ins, you are losing closer to $120 per winning all in.9 blinds? huh.. 8.4% = 1/12 hands, one round is 10 hands or 1,5bb, so you'd pay a little more than 1,5bb, which is ~$20, per all in.And, even if you lost $81 per all in to blinds(which is 5,4 rounds = 55hands), then after 8 all ins, you'd have ((1000-81)*2-81)*2-81)*2 ... = 215k so 8 all ins in a row is enough to win the whole 125k. This means that even if you went with only hands that are over 70% favorite, 8 all ins would be enough.Most importantly though, while your prior estimate based on .7 ^ 18 was conservative, when you only require an average of 70% winners you introduce a tremendous amount of variance, and a binomial distribution estimate is very inaccurate.It was .7^8 and that is quite accurate, as 8 all ins is enough, and your winning expectance really is 70% per all in(in the long run).Now, of course, variance is the key thing here:Assuming you have 20 buyins your risk of going broke before winning the 125k is (1-.7^8)^20= 30,5% (with only hands 88+ this is 7,15%)Which, of course, is very high.. But do note that you are trying to win 125k with only 20k bankroll.For comparison, say you only try to win 5 all ins in a row, or ~20k-30k.Then the risk of going broke is (1-.7^5)^20 = 2,5% (with 88+ this is 0,2%)Which is pretty good.

Link to post
Share on other sites

"9 blinds? huh.. 8.4% = 1/12 hands, one round is 10 hands or 1,5bb, so you'd pay a little more than 1,5bb, which is ~$20, per all in. "Unless I misunderstood something, we are playing HU, where do y0u get one round is 10 hands?"If you are only going with hands that are at least 70% favorite, then you'd actually be 77% favorite on average every all in. Then the calculation would go .77^8=12,4%, because some all ins you are 70% favorite, and some even 85% (AA), which in the long run equals to .77 with hand range of 88-AA. Play with www.pokerstove.com if you can't understand this. "I understand it just fine, thank you. The post that proposed the 70% said "Look, say i'm a 70% favorite to win every hand". You interpreted that to mean average 70%, I interpreted it to mean at least 70%. Your attitude here sux.Your .7^8 approximation is not "quite accurate". Besides being a binomial approximation to a more complex distribution, it ignores the effect of the blinds in the early hands, which are significant enough to cause an extra all in or two, which increases the risk of ruin.I ran 20 simulations (admittedly a small number, but my VBA looping isnt working so its more manual than it should be) and the average number of buyins was 19.9, and the 2ok bankroll was broken 11 out of the 20 sims. I'm stuck on a plane tomorrow and will run another 100 or so....or ill fix my looping and run a few thousand). The number of buyins ranged from 1 to 37. The sims use your 8.4% of hands, generates whether a hand is played, and if played whether its won or lost. It does use an average of 7.50 blind, rather than alternating between 5 and 10, but thats the only approximation.

Link to post
Share on other sites

Wow.This forum is full of ignorant people (and informed people as well, no offence).There are only three valid arguments against taking such a HU proposition (one where the blinds are small enough to allow skill): bankroll requirements, declining utility of money, and opportunity cost (and the first two are somewhat intertwined).1) Bankroll requirementsIf your entire bankroll consists of the buy-in, it's quite foolish to risk it all on a long shot (even if it has a positive expected value).2) Declining utility of moneyIf you have just enough money to live comfortably, it may hurt more to lose (and live uncomfortably) than to win (and continue to live comfortably), even if mathematically the win is, on average, larger.3) Opportunity costIf someone offers you a proposition with the stacks at $0.20 and $10, and the blinds at $0.01/$0.02, it would hardly be worth the time to play. Your time could be better spent working or playing poker at higher limits.

Link to post
Share on other sites
Unless I misunderstood something, we are playing HU, where do y0u get one round is 10 hands?Your .7^8 approximation is not "quite accurate". Besides being a binomial approximation to a more complex distribution, it ignores the effect of the blinds in the early hands, which are significant enough to cause an extra all in or two, which increases the risk of ruin.I ran 20 simulations (admittedly a small number, but my VBA looping isnt working so its more manual than it should be) and the average number of buyins was 19.9, and the 2ok bankroll was broken 11 out of the 20 sims. I'm stuck on a plane tomorrow and will run another 100 or so....or ill fix my looping and run a few thousand). The number of buyins ranged from 1 to 37. The sims use your 8.4% of hands, generates whether a hand is played, and if played whether its won or lost. It does use an average of 7.50 blind, rather than alternating between 5 and 10, but thats the only approximation.
Yeah, I thought we were talking about full ring, as dn was playing a full ring game. But my calculations are very accurate in full ring situation, they do take the blinds into account as I said earlier. Of course in heads up the variance increases significantly, as you'll lose lot more to blinds. This also means that the best way to play would probably be to only play avg 70% favorite hands early on and later tighten up to only 88+ pps, as the few percents makes huge difference because it's to the power of 8.Anyways, it should be clear by now that it is -ev to the 125k stack maniac to go all in blind every hand, and +ev to the other player..
Link to post
Share on other sites

This is such an easy concept... The only chips that are in play are determined by table stakes. If you have 1k and the other guy has 500, then it's the same as if you had 500 and he had 500. A +EV play with $500 will average winning money. A -EV play will lose money in the long run.That thing about busting a guy with a huge bankroll is fairly interesting. Try looking at Jim's Poker page. Look for doyle brunson vs. bill gates. You guys will be surprised by the real answers (done through math and big samples not by using rhetoric).

Link to post
Share on other sites
This is such an easy concept... The only chips that are in play are determined by table stakes. If you have 1k and the other guy has 500, then it's the same as if you had 500 and he had 500. A +EV play with $500 will average winning money. A -EV play will lose money in the long run.That thing about busting a guy with a huge bankroll is fairly interesting. Try looking at Jim's Poker page. Look for doyle brunson vs. bill gates. You guys will be surprised by the real answers (done through math and big samples not by using rhetoric).
1k vs 500 is different from 500 vs 500 in a freezeout format which is what is being discussed. Each individual hand is determined by the smaller stack, but the fact that both stacks are committed to play until one is busted gives the 1k player an advantage.
Link to post
Share on other sites
That thing about busting a guy with a huge bankroll is fairly interesting. Try looking at Jim's Poker page. Look for doyle brunson vs. bill gates. You guys will be surprised by the real answers (done through math and big samples not by using rhetoric).
The Brunson/Gates scenario is not the same as we are discussing, but closer to the Beal/Pro challenges because its fixed limit, not NL. It is akin to a biased coin flip with fixed stakes and no blinds, and assumes a fixed advantage in every round thus is very easily calculated without simulation from formulas or from binomial distribution tables. Cycracs math does the job just fine if its modified to fixed stakes instead of NL. The Brunson/Gates scenario on Jims poker pages is , in fact, extremely misleading because it ignores the size of the blinds relative to the initial stakes. Its conclusion that DB would only need about a 51/49 edge if his stake is 20 mill vs Gates 50 billion and a 1 million at a time buy in is nonsense unless the blinds are insignficant relative to the buyin. With big blinds similar to the Beal/Pro challenges that got as high as 4% of the buy in, the pros "short run" edge ( a lot more than 51/49) was pretty much eliminated and Beal would be a huge favorite to bust the pros eventually, and he wouldnt need 50 billion to do it.
Link to post
Share on other sites

The 51-49 edge that Jim is talking about means that one person will 51 times out of 100 after taking into account all the game conditions. It doesn't matter what the blinds are; the edge and the percentage of the bankroll at risk are what's important.

Link to post
Share on other sites

What you seem to be talking about is when someone sits down with 1k and another guy sits down with 100k, and they both decide to play until one loses all the their chips. In this scenario, obviously the guy with 1k will lose the vast majority of the time. However, if they are of equal skill and there is no rake, their EV is 0. If you look at this game in isolation, no one has an EV edge.However, the effect of this game on each player is different depending on how much of their bankroll is in play. This thread is about table stakes and how DN sat down with 125k in a 5-10 NL game. It is not about how DN and the other players both set down their entire bankroll at one game.Also, if some player had an edge in this type of game, he would expect to win money irrelavent of how much of his bankroll is at risk. He may be placing himself in an extremely risky position from his perspective, but his expected value will still be positive for this game. In the Beal/coporation games, if the corporation has a 1% edge with their whole bankroll at stake every hand, their expected value is still positive. It's just that every time Beal goes broke, the pros will go broke far more often. Don't confuse this with the expectation of the game.

Link to post
Share on other sites

The only reason that the pros would involve themselves in a game like this is due to the Kelly Criterion. They will not maximize their edge in the long run if they risk too great a fraction of their bankroll.

Link to post
Share on other sites
×
×
  • Create New...